Reduced Whole Life Cost by Engineering Design

Transcription

1 Reduced Whole Life Cost by Engineering Design Sin Sin Hsu 17 th July

2 Considering the Whole Life Cost From the start of CP5, WLC analysis is mandatory for all infrastructure projects WLC is not a new concept but it is one that we must apply more rigorously when developing projects in order to derive best value from our assets. 2

3 How can we reduce WLC by design? Various tools available: Vehicle Dynamics Simulations Wheel/rail interface issues Structural gauging VAMPIRE Track Ex (Network Rail s in house software) CAD and Finite Element Modelling (SolidWorks) System component stress analysis Assembly motion analysis 3

4 Case Studies 1. Alignment Design Optimisation Northern Hub Ordsall Lane Junction 2. Layout Design Optimisation Ordsall Chord Pts Component Design Optimisation Switches and Crossings

5 Case Study 1 Alignment Design Optimisation Collaboration Work to date: Northern Hub (Ordsall Lane, Ordsall Chord, Castlefield Junction, Poulton Le Fylde) Crossrail (Stockley Curves, Leigh Road) Routes (Cambria Spurs, Farnham) The philosophy is to Even out wear and RCF damage at the leading and trailing wheelsets Avoid high localised wear Maximise rail life 5

6 What do we know about low rail issues? Low rail heavy spalling Noise Low rail corrugation

7 Examples of Low Speed Sharp Radius Curves Site Curve Radius (m) Speed (mph) E (mm) E+D (mm) D (mm) D% of E Issues St Denys % Low rail (2010) corrugation 9 months after renewal Penarth Curve Liverpool St Stn Strood Fort Pitt Tunnel ?? % % Low rail RCF, plastic flow (7MGT, 50% freight) Low rail (2001) severe RCF, spalling and plastic flow Low rail (2001) severe RCF, spalling and plastic flow Vauxhall % Low rail (2008) corrugation and RCF 7

8 Low Speed, Sharp Radius Curves High applied cant: High angle of attack Prone to High Rail RCF High lateral forces Prone to Low Rail RCF

9 Low Speed, Sharp Radius Curves High applied cant: High angle of attack Prone to High Rail RCF High lateral forces High cant deficiency: Wheelsets aligned more radially Prone to Low Rail RCF Less lateral force and rail wear at the leading wheelset

10 Effect of Cant Deficiency 250 Wear due to the leading wheel on the low rail Tgamma Low Rail (N) % 50 D = Curve Radius (m) D = 110mm 10

11 Consider Actual Running Speed? Sharp curve at Stockley: 310m radius curve with 140mm cant and a linespeed of 50mph 150 mm (Exceptional) 104 mm (50mph) 16 mm (40mph) -52 mm (30mph) Tgamma (N) % High Rail Gauge corner Existing line speed - 50mph 40mph 30mph Max D If vehicle is travelling at 30mph, wear at the low could be 40% higher! Distance (m) - 45% 200 Cant deficiency should be maximised in curves to reduce wear at the low rail Low Rail Distance (m)

12 Current Policy on Cant Deficiency TRK/2049 states that: B.2.4: Curving Design Values Guidance on Circular Curves Cant & Cant Deficiency All curves should be designed to operate with some cant deficiency to aid steering. Curves should be designed so that the cant deficiency (for conventional trains) does not normally exceed 60% of the applied cant on jointed track or 73% of the applied cant in continuously welded track. Exceptionally this may be increased to 73.3% for jointed track and in excess of 100% for continuously welded track. Values in excess of 100% in CWR will require the approval of the Head of Track Engineering at Network Rail HQ. 12

13 What does it mean? Typical applied cant on low speed curves is 70mm, this limits the cant deficiency to 73%, i.e. D = 50mm; exceptionally, D can be 70mm If D greater than 100% is required, then needs approval from PHoT So, what is the ratio for a flat curve with no cant? Note: The requirement is the approval from PHoT, therefore not a TNC! 13

14 Design Limits Infinity for flat curves Less Low Rail Damage Example 300m Radius Curve 15mph 20mph 25mph 30mph 35mph %D of E mph 45mph 50mph For low speed, sharp radius curve, it is recommended that D is at least 70mm Limits set in B Applied Cant (mm) 14

15 Ordsall Lane Proposed Option 20 Drawing No: E-ETR-DRG-PBL A04 Sections considered: 1. Down Chat Moss 2. Up Chat Moss 3. Down Bolton (COL, OLW) 4. Up Bolton (COL, OLW) 5. NIRU Siding 6. NIRU Connection 15

16 Main Simulation Output Main Output Description Values Wear, T (pronounced T gamma) Energy at contact patch wear rate of rails Moderate to Severe > 200 N 65 N < Mild to Moderate < 200 N Negligible to Mild < 65 N RCF Damage Index 1 is the maximum damage for 260 grade steel Moderate to Severe > < Mild to Moderate < 0.8 Negligible to Mild < 0.4 Note: The traffic light colour code is used to present predicted severity of damage 16

17 Example Results Down Chat Moss Element No Speed Radius Eq Cant (E+D) Ea D mph m mm mm mm C C C C C C C C C C C C C

18 Example Results Down Chat Moss Wear RCF Element No Speed Radius Eq Cant (E+D) Ea D High rail crown T High rail gauge corner T Low rail T High rail crown RCF High rail gauge corner RCF Low rail RCF Comments mph m mm mm mm N N N C C C C C CEN 56 DV split equal Junction C C Existing track C Existing track C Existing track C Existing track C EVs, 2603 pts C C EVs, 6531 pts Proposed design changes and/or maintenance regimes for high wear/ RCF regions 18

19 Recommendations Maximise cant deficiency (above 70mm) Use premium rail steels Lubricate (curves <500m radius) Result: Easier construction (lower cant) Reduced wear and RCF and therefore less maintenance and lower whole life cost 19

20 Case Study 2 Layout Design Optimisation AIP proposed Design Through: R=350m, E=35mm, D=44mm Turnout: R=227m, E=35mm, D=86mm Alternate Solution Through: R=306m, E=0mm, D=90mm Turnout: R=583m, E=0mm, D=47mm 20

21 Predicted Rail Life Section Route Proposed Maximum EMGTPA Predicted Rail Life, Years Existing Design Alternate Design Through (R=350m, E=35mm, D=44mm) Turnout (R=227m, E=35mm, D=86mm) Through (R=306m, E=0mm, D=90mm) Turnout (R=583m, E=0mm, D=47mm)

22 Recommendations Like for like renewals may not always give the best WLC Increase cant deficiency where possible Increase curve radius whenever possible Zero cant if possible 22

23 Result Easier construction (no applied cant on the switches) Less reliability issues with pushing points against gravitational forces from the canted track Reduced rail wear and therefore less maintenance and lower whole life cost Reduced wear at the switch planing region and therefore less risk of train derailment 23

24 Case Study 3 Component Optimisation 24

25 Available Crossings Product Type Crossing type Crossing Angle Cost Range Fabricated Acute 1 in 4 to 1 in 21 3,600 to 5,300 Obtuse 1 in 4.75 to 1 in 8 3,800 to 4,400 Part-welded Acute 1 in 4 to 1 in 20 8,500 to 12, lb CCB 113lb Cast Monobloc NR60 Cast Monobloc Acute Acute Acute 1 in 8 to 1 in in 8 to 1 in in 7 to 1 in ,900 to 18,500 13,800 to 26,300 13,500 to 36,500 Obtuse Obtuse Obtuse 1 in 5.5 to 1 in 8 1 in 5.5 to 1 in 8 1 in 5.25 to 1 in ,700 to 18,900 13,500 to 20,900 23,800 to 24,000 25

26 Available Switches Product Type Bearer Type Switch Length Cost Range 113A Full Depth Switches 113A Shallow Depth Timber Timber AV - GV AV - GV 7,700 to 22,900 12,000 to 32,800 Concrete Concrete AV - GV AV - GV 7,700 to 21,800 12,000 to 30,400 NR60 inclined Concrete C - H 17,400 to 45,600 26

27 Cost of S&C Failures Cost of Delay Minutes per Year Network Rail Route Point Operating Grand Total Per Year Ironwork Equipment Anglia 3,489, , ,170, East Midlands 795, , ,310, Kent 3,433, , ,864, LNE 3,496, ,713, ,210, LNW 6,787, ,013, ,800, Scotland 1,026, , ,298, Sussex 2,768, ,292, ,060, Wales 473, , , Wessex 1,435, , ,339, Western 5,032, ,374, ,407, Grand Total 28,738, ,272, ,011, This is the real cost! 27

28 Switches where should we focus? Similar flexure switches that may at any time or under any circumstances be used in the facing direction Turnout radius of 300m and below on a through main line radius of 1000m and below Generally these are switches with lengths of E and shorter (inclusive of NR60, RT60 and CEN56E1) High risk switches require 13 weekly inspection 28

29 High Risk Switches Edinburgh 2011 Avoid similar flexure turnouts Increase turnout radius where possible Zero cant if possible Shrewsbury 2011 Waterloo x Glasgow 2007 Victoria 2007 Waterloo x

30 Recommendations Avoid similar flexure turnouts Increase turnout radius where possible Zero cant if possible 30

31 Crossings where should we focus? 1. Main Lines (Track ID 2100 and 1100) over 60% of RDMS defects High tonnage (> 16MGTPA) High Speed (> 70mph) 2. Wide angle crossings (sharper than 1 in 6.5) due to large dip angles 3. Crossings on high rail of high cant deficiency curves 4. Crossings on low rail of low cant deficiency curves freight routes 5. NR60/RT60 31

32 Wide Crossing Angles Sharp angle Wheel transfer near or at nose Large dip angle and therefore impact forces Flat angle More support from the wing rail before wheel transfers to crossing nose Less damage to nose for a given speed Higher vehicle speed increases the dynamic forces 21 - Jul 32 -

33 Examples of wheel transfer 1 in 10 Wing and nose overlap 1 in in 4.5 Flat wing rail Inclined wing rail 1 in 9.5 Inclined wing rail Inclined wing rail 21 - Jul 33 -

34 Issues with NR60/RT60 Common Crossings 1. Large dynamic forces at wheel transfer area 2. Casting foot cracking Cast monobloc type on concrete bearers Crack at crossing foot at the wheel transfer area 3. Wing rail cracking? Crack at wing rail at the wheel transfer area 34

35 Casting Foot Cracking Cast monobloc foot flange design Casting base design Cast monobloc Part welded 35

36 Issues with NR60/RT60 Obtuse Crossings Casting foot cracking Cast monobloc foot flange design 36

37 Issues with NR60/RT60 Obtuse Crossings Casting foot cracking Cast monobloc foot flange design Casting base design / Bearer spacing 37

38 Obtuse Crossing Base Design No bearer support under the smallest section? 38

39 Bearer Position for RT60? Great West (Basingstoke) BBMC 1 in 4.5 Feltham Outreau 1 in 6.25 Weybridge EA 1 in 7.5 Vauxhall Outreau 1 in

40 Bearer Position? Great West (Basingstoke) BBMC 1 in 4.5 Weybridge EA 1 in 7.5 Feltham Outreau 1 in 6.25 Vauxhall Outreau 1 in No support under knuckle and crossing nose No support under knuckle Bearer under knuckle

41 Continuous Foot Flange Second moment of area (I) geometrical strength of a section to withstand loading Continuous foot flange Increases I xx by 44% Reduces the height of the neutral axis (less tension at the foot) Resulting in the overall foot stress reduction of up to 40% My I xx 41

42 Closing up an open section? Closing up an open section Increases I xx by 22% Reduces the height of the neutral axis Resulting in the overall foot stress reduction of up to 20% 42

43 Both changes? With both the continuous foot flange and closed up section, the total reduction in foot stress could be up to 55% 43

44 Bearer Spacing Historically, 113lb obtuse crossings have shorter span at the wheel transfer area 1 in 5.5 crossing with 419mm bearer spacing at the knuckle E.g for 1 in 5.5 angle, spacing is 419mm at knuckle Philosophy appears to be bearer at the knuckle Bearers closer together at the knuckle 44

45 Implications? RT60/NR60 standardised 650mm Great West Junction 1 in 4.5 obtuse crossing has been replaced twice in 2 years! NR56 obtuses also designed with same spacing! Performance could potentially be worse due to smaller rail section 45

46 Effect of Reducing Bearer Spacing P Reduce the spacing to 400mm, difference in bending moment and therefore foot stress is 40%! L M = (PL)/4 at mid point My I xx Stress at casting foot increases with bending moment! 46

47 Proposed Crossing Improvement

48 Recommendations Flatter crossing angles for main lines (including diamonds) Inclined wing rail design for all common crossings Avoid putting crossings on curves, whenever possible Avoid two levelling Specify all crossings to have continuous foot flanges 1 in 20 In future, cast obtuse crossings to have smaller bearer spacings at the wheel transfer area 48

49 Result Better performance of the crossings by reduced impact forces / fatigue loading at the wheel transfer area leading to longer crossing life Reduced voiding and associated component deterioration Less maintenance activities including weld repairs and possession time Reduced risk of cracked crossings and therefore ESRs and associated delay minutes 49

50 Conclusions The most effective way of reducing WLC is at the earliest stage of the design process. Track design can be improved for easier installation and reduced maintenance and therefore lower whole life cost Good engagement with suppliers/contractors to ensure better quality of product and consistent production processes to improve asset performance Safety by design, reducing maintenance liability and therefore less staff on track! 50